Anirudha Ramesh

Anirudha Ramesh

Graduate Student at Robotics Institute, Carnegie Mellon University.

Posts by Collection

portfolio

publications

Multi-object Monocular SLAM for Dynamic Environments

Published in IEEE Intelligent Vehicles Symposium 2020 (IV), 2020

We present a metric scale absolute novel solution for simultaneous localization and mapping (SLAM) in the challenging scenario of autonomous driving platforms equipped with only a monocular camera

BirdSLAM: Monocular Multibody SLAM in Bird’s-Eye View

Published in VISIGRAPP (VISAPP), 2021

We present a metric scale absolute novel solution for simultaneous localization and mapping (SLAM) in the challenging scenario of autonomous driving platforms equipped with only a monocular camera in an orthographic Bird’s Eye View (BEV) which is most suitable for down-stream.

Probabilistic Collision Avoidance For Multiple Robots: A Closed Form PDF Approach

Published in IEEE Intelligent Vehicles Symposium (IV), 2021

We propose a novel method for reactive multiagent collision avoidance by characterizing the longitudinal and lateral intent uncertainty along a trajectory as a closed-form probability density function. We do so by introducing it into the Time Scaled Collision Cone (TSCC) approach.

What Ails One-Shot Image Segmentation: A Data Perspective

Published in Datasets and Benchmarks, NeurIPS, 2021

Identifying and tackling implicit biases in few shot learning systems (few shot segmentation) from a data perspective, which leads to general performance improvements for networks across the state of the art.

Intelligent Railway Capacity and Traffic Management Using Multi-Agent Deep Reinforcement Learning

Published in IEEE Intelligence Transportation Systems Society (ITSC), 2024

A fundamental centerpiece of future digitized railway network operations is automated and optimized planning and dispatching. The sector initiative “Digitale Schiene Deutschland” (DSD) develops a holistic and intelligent Capacity & Traffic Management System (CTMS) that can automatically plan and continuously optimize railway traffic at scale. Both, planning and dispatching tasks, are highly complex and, today, require human expertise and oversight. Our main contribution is a multi-agent deep reinforcement learning approach at the core of the envisioned CTMS, which learns from interaction with a realistic, microscopic railway simulation. Our results demonstrate that the proposed approach flexibly solves planning and re-scheduling tasks in the realistic setting of a medium-sized part of the German railway network. It exhibits response times and scaling properties that make it a promising candidate for future applications in railway operations at scale.

talks

teaching